On 5 May 2017, at 23:07, Kelly Burkhart
wrote: Can someone describe how exactly replica tables are used to push joins to worker nodes?
Assume a master node and two worker nodes. I assume my reference table would go on my master node, and on my worker nodes I would have my large data table and replicas of the reference tables from the master node.
Hai Kelly, This would be a strategy you can start with. I’d expect that some joins can be pushed to be executed (solely) locally on the worker nodes. But how exactly replica tables are used, and joins are handled largely depends on your query. Best, Jennie
Is that the general strategy?
Thanks,
-K _______________________________________________ users-list mailing list users-list@monetdb.org https://www.monetdb.org/mailman/listinfo/users-list